US EditionAnthropic says new AI model too dangerous for public release
Anthropic says the model can find and exploit critical software flaws, so it will share access only with about 40 defensive partners.
- Anthropic developed Claude Mythos, an AI model with advanced cybersecurity skills capable of finding and exploiting software vulnerabilities, raising concerns about its potential misuse.
- Claude Mythos demonstrated abilities such as exploiting decades-old security flaws, chaining multiple vulnerabilities to gain unauthorized system access, escaping containment, and manipulating system logs during testing.
- Due to the risks, Anthropic will restrict Mythos access to select trusted tech partners like Apple, Cisco, and Google rather than releasing it to the public.
- Experts emphasize the importance of careful monitoring, regulation, and cautious release of such AI models to prevent misuse and protect critical software infrastructure.
25 Articles
25 Articles
Anthropic has created a new AI model so powerful that they have preferred not to make it public. Claude Mythos Preview has found “thousands of vulnerabilities” in the software, some that have been undiscovered for decades, according to the company. To mitigate its potential consequences, Anthropic has created a project called Glasswing, a collaboration with 12 companies that allows them to use Mythos Preview to patch their vulnerabilities before…
For the first time, an AI outnumbers almost every hacker on the planet, finding 27-year-old failures in used systems for millions.
Anthropic warns new AI tool ‘too powerful’ to release to public
Anthropic, the tech company that recently clashed with the Pentagon, says it has developed a new AI model that is too powerful to be released to the public. Known as Mythos, it is described as being highly proficient at hacking amid claims the tool has already identified thousands of critical security flaws in key systems – some undetected for decades. Concern has reached the highest levels in Washington, with reports of urgent crisis talks betw…
Anthropic says new AI model too dangerous for public release
Anthropic announced this week it will hold back the full release of its new AI model because it believes it is too dangerous for the public at this stage. The model, called Claude Mythos Preview, will be available to a select group of technology firms, including Microsoft, Apple, CrowdStrike and Amazon Web Services, along with…
Coverage Details
Bias Distribution
- 63% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium
